The myeloid leukemias are a major cause of human mortality and morbidity. The search for genetic and epigenetic origins of the myelogenous leukemias has been advanced by many technologies, so we now have an excellent list of leukemia protooncogenes, leukemogenic events, and the leukemias that often harbor them, but we have little idea of the order of events that occur in the progression from normal through preleukemic stages to frank acute leukemias, and additionally, the possibility that the leukemic progression goes through known myeloid stem and progenitors to give rise to various stages of leukemia stem cells (LSC). We proposed, that the only cells that persist through the preleukemic clonal progression and therefore could accumulate the several independent, rare events are self-renewing stem cells, and in leukemia, HSC, unless the first event allowed a non self-renewing progenitor to gain self-renewal capacity. If this hypothesis is correct, one should be able to isolate LSC from all leukemic stages, and in the various malignant myeloproliferative disorders. We have developed several mouse models of leukemia and we propose to identify the LSC in each of them. As there are many forms of human myeloproliferative disorders (MPDs) and leukemias, we feel that it is appropriate and advantageous to investigate multiple mouse models as each has unique properties and parallels to human diseases. In each of these models, we will use rigorously purified populations in transplantation assays, cDNA microarray analysis, quantitative PCR and additional highly sensitive assays to identify and characterize the LSC associated with each of these models. We propose to investigate the potential protooncogenes our studies reveal in both mouse and human myelogenous leukemias. It is the goal of these studies to provide further confirmation of the cancer/leukemia stem cell hypothesis, and to provide insight into the leukemogenic process and potential therapies.